[Quantlib-users] Hull White model volatility

[Hristo R. <hri...@gm...>]

Hi,

I'm trying to value callable bonds with Quantlib for Python using Hull
White model and TreeCallable engine.

 model = ql.HullWhite(tts_handle, mean_reversion, volatility)
 engine = ql.TreeCallableFixedRateBondEngine(model, grid_points)

I saw that default parameters for mean reversion and volatility are  a=0.1,
sigma=0.01. My question is in what terms is volatility/sigma expected as
input? Is it in terms of absolute basis points value or in terms of percent
of some rate?
Does sigma of 0.01 mean standard deviation per year of 100bps?  For
example  let say rate is 0.04(4%) and at valuation point it will go up/down
with 100bps in absolute value to 0.05 (5%) / 0.03 (3%). Or rate will
increase / decrease by sigma(0.01) percent of the rate 0.04 +/- 0.04*0.01 =
0.04 (4%) +/- 0.0004 (4bps)?